Laser annealing for two-qubit gate error reduction in fixed frequency processor architectures

Fixed-frequency transmon qubits are attractive due to noise immunity and high coherence. However, as quantum processors are scaled, frequency crowding resulting from level degeneracies reduces two-qubit gate fidelity [1]. We implement a scalable laser-based frequency trimming technique that has been implemented on processors at the 65-qubit scale, enabling low-error gates with average fidelities approaching 99%. We discuss ultimate tuning precision of the method and its consequences for processor scaling.

[1] J. Hertzberg, S. Rosenblatt, J. Chavez, E. Magesan, J. Smolin, J. B. Yau, V. Adiga, M. Brink, E. Zhang, J. Orcutt, J. Chow, “Effects of qubit frequency crowding on scalable quantum processors.” Bulletin of the American Physical Society. 65, Mar 5, 2020 *We acknowledge support for modeling work under IARPA under Contract No. W911NF-16-0114